An integrated molecular diagnostic report for heart transplant biopsies using an ensemble of diagnostic algorithms

Parkes, MD; Aliabadi, AZ; Cadeiras, M; Crespo-Leiro, MG; Deng, M; Depasquale, EC; Goekler, J; Kim, DH; Kobashigawa, J; Loupy, A; Macdonald, P; Potena, L; Zuckermann, A; Halloran, PF

An integrated molecular diagnostic report for heart transplant biopsies using an ensemble of diagnostic algorithms

Parkes, MD Aliabadi, AZ Cadeiras, M Crespo-Leiro, MG Deng, M Depasquale, EC Goekler, J Kim, DH Kobashigawa, J Loupy, A Macdonald, P

Potena, L Zuckermann, A Halloran, PF

Permalink

Publication Type:: Journal Article
Citation:: Journal of Heart and Lung Transplantation, 2019, 38 (6), pp. 636 - 646
Issue Date:: 2019-06-01

Closed Access

	Filename	Description	Size
	Parkes_Integrated.pdf	Published Version	861.55 kB	Adobe PDF	View/Open

Copyright Clearance Process

Recently Added
In Progress
Closed Access

This item is closed access and not available.

Full metadata record

Field	Value	Language
dc.contributor.author	Parkes, MD	en_US
dc.contributor.author	Aliabadi, AZ	en_US
dc.contributor.author	Cadeiras, M	en_US
dc.contributor.author	Crespo-Leiro, MG	en_US
dc.contributor.author	Deng, M	en_US
dc.contributor.author	Depasquale, EC	en_US
dc.contributor.author	Goekler, J	en_US
dc.contributor.author	Kim, DH	en_US
dc.contributor.author	Kobashigawa, J	en_US
dc.contributor.author	Loupy, A	en_US
dc.contributor.author	Macdonald, P https://orcid.org/0000-0001-5378-2825	en_US
dc.contributor.author	Potena, L	en_US
dc.contributor.author	Zuckermann, A	en_US
dc.contributor.author	Halloran, PF	en_US
dc.date.accessioned	2020-04-16T02:20:38Z
dc.date.available	2019-01-31	en_US
dc.date.available	2020-04-16T02:20:38Z
dc.date.issued	2019-06-01	en_US
dc.identifier.citation	Journal of Heart and Lung Transplantation, 2019, 38 (6), pp. 636 - 646	en_US
dc.identifier.issn	1053-2498	en_US
dc.identifier.uri	http://hdl.handle.net/10453/140012
dc.description.abstract	© 2019 International Society for Heart and Lung Transplantation BACKGROUND: We previously reported a microarray-based diagnostic system for heart transplant endomyocardial biopsies (EMBs), using either 3-archetype (3AA)or 4-archetype (4AA)unsupervised algorithms to estimate rejection. In the present study we examined the stability of machine-learning algorithms in new biopsies, compared 3AA vs 4AA algorithms, assessed supervised binary classifiers trained on histologic or molecular diagnoses, created a report combining many scores into an ensemble of estimates, and examined possible automated sign-outs. METHODS: We studied 889 EMBs from 454 transplant recipients at 8 centers: the initial cohort (N = 331)and a new cohort (N = 558). Published 3AA algorithms derived in Cohort 331 were tested in Cohort 558, the 3AA and 4AA models were compared, and supervised binary classifiers were created. RESULTS: A‘lgorithms derived in Cohort 331 performed similarly in new biopsies despite differences in case mix. In the combined cohort, the 4AA model, including a parenchymal injury score, retained correlations with histologic rejection and DSA similar to the 3AA model. Supervised molecular classifiers predicted molecular rejection (areas under the curve [AUCs]>0.87)better than histologic rejection (AUCs <0.78), even when trained on histology diagnoses. A report incorporating many AA and binary classifier scores interpreted by 1 expert showed highly significant agreement with histology (p < 0.001), but with many discrepancies, as expected from the known noise in histology. An automated random forest score closely predicted expert diagnoses, confirming potential for automated signouts. CONCLUSIONS: Molecular algorithms are stable in new populations and can be assembled into an ensemble that combines many supervised and unsupervised estimates of the molecular disease states.	en_US
dc.relation.ispartof	Journal of Heart and Lung Transplantation	en_US
dc.relation.isbasedon	10.1016/j.healun.2019.01.1318	en_US
dc.subject.classification	Surgery	en_US
dc.title	An integrated molecular diagnostic report for heart transplant biopsies using an ensemble of diagnostic algorithms	en_US
dc.type	Journal Article
utslib.citation.volume	6	en_US
utslib.citation.volume	38	en_US
utslib.for	1102 Cardiorespiratory Medicine and Haematology	en_US
pubs.embargo.period	Not known	en_US
pubs.organisational-group	/University of Technology Sydney
pubs.organisational-group	/University of Technology Sydney/Faculty of Health
utslib.copyright.status	closed_access
pubs.issue	6	en_US
pubs.publication-status	Published	en_US
pubs.volume	38	en_US

Abstract:

© 2019 International Society for Heart and Lung Transplantation BACKGROUND: We previously reported a microarray-based diagnostic system for heart transplant endomyocardial biopsies (EMBs), using either 3-archetype (3AA)or 4-archetype (4AA)unsupervised algorithms to estimate rejection. In the present study we examined the stability of machine-learning algorithms in new biopsies, compared 3AA vs 4AA algorithms, assessed supervised binary classifiers trained on histologic or molecular diagnoses, created a report combining many scores into an ensemble of estimates, and examined possible automated sign-outs. METHODS: We studied 889 EMBs from 454 transplant recipients at 8 centers: the initial cohort (N = 331)and a new cohort (N = 558). Published 3AA algorithms derived in Cohort 331 were tested in Cohort 558, the 3AA and 4AA models were compared, and supervised binary classifiers were created. RESULTS: A‘lgorithms derived in Cohort 331 performed similarly in new biopsies despite differences in case mix. In the combined cohort, the 4AA model, including a parenchymal injury score, retained correlations with histologic rejection and DSA similar to the 3AA model. Supervised molecular classifiers predicted molecular rejection (areas under the curve [AUCs]>0.87)better than histologic rejection (AUCs <0.78), even when trained on histology diagnoses. A report incorporating many AA and binary classifier scores interpreted by 1 expert showed highly significant agreement with histology (p < 0.001), but with many discrepancies, as expected from the known noise in histology. An automated random forest score closely predicted expert diagnoses, confirming potential for automated signouts. CONCLUSIONS: Molecular algorithms are stable in new populations and can be assembled into an ensemble that combines many supervised and unsupervised estimates of the molecular disease states.

Please use this identifier to cite or link to this item:

http://hdl.handle.net/10453/140012